To The Moon

AVGO's latest numbers create one of the more interesting debates in AI semiconductors right now.

Broadcom's AI semiconductor revenue reached $16.7B in Q3, up an extraordinary 221% YoY and 54% sequentially. Management expects that figure to accelerate again to approximately $21.7B in Q4.

Yet AVGO is around $355, well below its June peak.

Why?

Because the market isn't questioning whether Broadcom can sell AI chips anymore.

It is questioning how profitable each additional dollar of AI revenue will be.

That distinction determines whether $355 is an opportunity or a value trap.

The $16.7B Number Is Bigger Than It Looks

AI semiconductor revenue now represents roughly 56% of Broadcom's total revenue.

Custom accelerators are no longer an interesting side business. They are rapidly becoming the centre of the company.

Broadcom's six XPU customers are accelerating deployments. XPU shipments increased more than 3.5x YoY, while AI networking revenue increased more than 2.5x.

And Q4 guidance suggests there is no immediate slowdown.

AI semiconductor revenue:

Q1: $8.4B

Q2: $10.8B

Q3: $16.7B

Q4 guidance: $21.7B

That is an extraordinary curve.

But there is another number investors need to watch.

75%.

That was Q3 gross margin, down 210 basis points sequentially.

Why?

The AI semiconductor business is becoming a larger percentage of Broadcom's revenue mix.

Ironically, the faster Broadcom's AI business grows, the more pressure it can put on consolidated gross margin.

That is the paradox behind the current selloff.

But I Think the Market May Be Focusing on the Wrong Margin

Gross margin fell.

Operating economics didn't collapse.

Broadcom generated $20.1B of non-GAAP operating income, up 92% YoY.

Operating margin reached approximately 68%, actually improving YoY because revenue is growing much faster than operating expenses.

And free cash flow reached a remarkable $13.7B, equivalent to about 46% of revenue.

That changes how I interpret the margin debate.

If gross margin falls because a lower-margin business is exploding in size, but operating profit, EPS and free cash flow continue expanding rapidly, that isn't necessarily margin deterioration.

It can be mix evolution.

This is the number I would watch next:

Can operating leverage continue outrunning gross-margin compression?

If yes, I think the market eventually stops punishing AVGO for its AI mix.

Anthropic Makes This More Interesting

Broadcom's custom-silicon opportunity is also becoming more deeply embedded in the AI ecosystem.

Anthropic's IPO filing revealed financing arrangements under which Broadcom could provide up to $42B to help fund infrastructure using Broadcom-designed custom accelerators.

That introduces another layer of risk because financing customers to consume your technology deserves scrutiny.

But it also tells us something important about demand.

The bottleneck isn't necessarily finding customers for AI compute.

It is financing enough infrastructure to satisfy them.

Broadcom is increasingly sitting in the middle of that buildout.

And custom ASICs are becoming strategically important because hyperscalers don't necessarily want every AI workload running on expensive general-purpose GPUs forever.

That gives Broadcom a potentially enormous lane beside Nvidia rather than requiring it to replace Nvidia.

My AVGO Pick Levels

At approximately $355, I am interested, but I wouldn't deploy everything immediately.

$345-$355: First Pick

This is where I would begin accumulating.

AVGO has already corrected substantially from its peak while the underlying AI revenue trajectory has accelerated.

$325-$340: Preferred Pick

This is my favourite risk/reward zone.

If AVGO reaches here because of valuation compression or broader semiconductor weakness rather than deterioration in AI orders, I would increase my position.

$300-$320: Strong Pick

At this point I would become considerably more aggressive, assuming AI revenue guidance and cash generation remain intact.

$370-$380: First Confirmation

A sustained reclaim would tell me buyers are beginning to absorb the current valuation concerns.

$400-$420: Re-rating Zone

Above $400, I think the market begins treating the margin issue as manageable rather than structural.

$450-$480: Bullish Extension

This requires more than another big AI revenue number.

I would want evidence that Broadcom can simultaneously sustain exceptional XPU growth, protect operating margins and convert that growth into cash.

Below $300: Thesis Check

I wouldn't automatically average down.

At that point I would investigate whether AI orders, customer concentration, financing exposure or profitability expectations had materially changed.

My Short-Term Prediction

For the next 1-2 weeks, my base case is bottom-building rather than an immediate vertical recovery.

I assign roughly a 50% probability that AVGO trades around $340-$370 while investors digest the margin debate before attempting another move higher.

There is approximately a 30% probability that semiconductor weakness pushes AVGO toward $325-$340. That would actually be my preferred buying scenario.

And I assign around 20% to a stronger momentum recovery above $380, potentially opening $400 relatively quickly.

The catalyst I want isn't another headline saying AI demand is enormous.

We already know that.

I want evidence that profit dollars are scaling almost as impressively as revenue dollars.

What Would Make Me Wrong?

There are three risks I would watch closely.

First, gross-margin compression becomes progressively worse while operating leverage stops compensating for it.

Second, Broadcom increasingly has to finance customers to sustain AI infrastructure demand, creating greater balance-sheet or counterparty exposure.

Third, hyperscaler ASIC programmes become more competitive and Broadcom loses pricing power.

Those would change the thesis.

A 200-basis-point gross-margin decline by itself doesn't.

My Call

The market sees:

$16.7B AI revenue + falling gross margin = problem.

I see a more nuanced equation:

$16.7B AI revenue + 221% growth + rising operating leverage + $13.7B free cash flow = a margin question worth investigating, not an AI thesis that is breaking.

My map:

$345-$355: First Pick

$325-$340: Preferred Pick

$300-$320: Strong Pick

Above $380: Confirmation

$400-$420: Re-rating

Below $300: Thesis Check

Broadcom doesn't need custom ASICs to have software-like gross margins.

It needs the profit dollars created by AI to grow faster than the margin percentage falls.

So far, that is exactly what is happening.

And that is why I think the most important Broadcom number next quarter won't simply be AI revenue.

It will be how much operating profit that extraordinary AI growth produces.

I am not a financial advisor. Trade wisely, Comrades!

What's your take, Tigers? 🐯 Would you buy AVGO around $355, wait for $330, or does the falling gross margin worry you more than the AI growth excites you?

If you found this useful, give it a 👍, repost, and follow me for my next market read! 📈

If you find this post interesting, give it wings! 🪽 Repost and share the insights.

Do consider “Follow me” and get firsthand read of my daily new post. Thank you. @Tiger_SG  @Tiger_CashBoostAccount  @TigerEvents  @TigerStars  @TigerPM  @Daily_Discussion